Angazenn aeffe27b30 [Perf]set moe w2_weight default to be nz (#2842) ...

### What this PR does / why we need it?

This PR sets the default format of GMM w2_weight in w8a8_dynamic to be
NZ to improve performance.

### Does this PR introduce _any_ user-facing change?

No.

### How was this patch tested?


- vLLM version: main
- vLLM main:
e40827280b

---------

Signed-off-by: Angazenn <supperccell@163.com>

2025-09-11 21:40:54 +08:00

..

__init__.py

[Core] Cherry pick from 0.7.1 to keep the main code newest (#127)

2025-02-21 17:07:37 +08:00

quant_config.py

support qwen25 vl w8a8 quantization (#2778)

2025-09-11 16:40:51 +08:00

utils.py

[2/N][Refactor][Quantization] clean quantization patch (#2785)

2025-09-08 17:31:53 +08:00

w4a8_dynamic.py

[main] [refactor] refactor common_fused_moe.py (#2706)

2025-09-08 20:09:50 +08:00

w8a8_dynamic.py

[Perf]set moe w2_weight default to be nz (#2842)

2025-09-11 21:40:54 +08:00

w8a8.py

[main] [refactor] refactor common_fused_moe.py (#2706)

2025-09-08 20:09:50 +08:00